Perlecan, originally named heparan sulfate proteoglycan, is now known to be an important component of all basement membranes (along with collagen type IV and laminin) and is thought to play a role in wound healing and angiogenesis. Perlecan consists of three heparan sulfate side chains linked to a large core protein of approximately 450 kDa (32, 36). This sequence has a single open reading frame of at least 3,707 amino acids that encodes for a protein of 396-466 kDa. Sequence analysis of the deduced sequences show the protein consists of five different domains, most of which contain internal repeats. Domain I contains a start methionine followed by a typical signal transfer sequence and a unique segment of 172 amino acids that contains the three probable sites of heparan sulfate attachment (of the amino acid sequence SGD). Domain II contains four strictly conserved cysteine-rich and acidic amino acid repeats that are very similar to those found in the LDL receptor and proteins such as megalin/GP330. Domain III consists of cysteine-rich and globular regions, both of which show similarity to those in the short arm of the laminin A chain. Domain IV contains 14 repeats of the immunoglobulin superfamily that are most highly similar to the immunoglobulin-like repeats in the neural cell adhesion molecule, and it appears domain IV has the capacity for differential splicing. Recombinant domain IV of perlecan binds to nidogens, laminin-nidogen complex, fibronectin, fibulin-2 and heparin (24). Domain V is the COOH-terminal domain and contains three repeats with similarity to the laminin A chain G domain. The repeats are separated by epidermal growth factor-like regions not found in the laminin A chain. Perlecan domain V is considered important in the supramolecular assembly of, and cell connections to, basement membranes (5).
The primary structural data agree with the appearance of the molecule in the electron microscope as a series of globules separated by rods, or “beads on a string.” Therefore, the name perlecan was adopted for this molecule (36). In summary, the variety of domains in perlecan suggests multiple interactions with other molecules, and each domain of native perlecan has the potential for separate functional activities related to wound healing and/or angiogenesis.
Expression of Perlecan
Native perlecan has been purified from the mouse Englebroth Holm Swarm (EHS) tumor (22), porcine kidneys (16), bovine kidneys (21), from the conditioned medium of bovine aortic endothelial cells (39), and from the extracellular matrix of cultured human fetal lung fibroblasts (2). Other sources of perlecan may also be available and the above list is not meant to be inclusive. The individual domains of perlecan have also been individually expressed in native configuration from bacterial hosts (33) and eukaryotic hosts (8, 10, 20). The obvious domain structure and limited susceptibility to proteolysis (11) contribute to the ability to produce each of the perlecan domains individually. For example, domain 1 has been cloned and expressed alone (10, 19), and with portions of domain II (20). It is known that sequences in the protein core affect the amount and type of glycosaminoglycan (such as heparan sulfate) which is attached when perlecan is produced in a eukaryotic host. However, the biological significance of the different glycosylation patterns remains unknown.
In situ hybridization studies and immunoenzymatic studies show a close association of perlecan with a variety of cells involved in the assembly of basement membranes, in addition to being localized within the stromal elements of various connective tissues. Perlecan has been demonstrated in periodontal ligament fibroblasts (29). Perlecan has also been detected in the basement membranes of human tissues including pituitary gland, skin, breast, thymus, prostate, colon, liver, pancreas, spleen, heart, and lung. All vascular basement membranes reported to be tested contained perlecan. In addition, sinusoidal vessels of liver, spleen, lymph nodes, and pituitary gland expressed high levels of perlecan in the subendothelial region. In situ hybridization, using as probe human cDNA-encoding Domain III, localized perlecan mRNA to specific cell types within the tissues and demonstrated that in skin, perlecan appears to be synthesized exclusively by connective tissue cells in the dermal layer (32, 33). Perlecan is also highly expressed in human bone marrow (28) and in synovium (13). An immunohistochemical study confirmed the location of perlecan on the apical cell surface of endothelial cells, and additionally as a dense fibrillar network surrounding the cells. In this context, the binding of thromobospondin 1 to the apical surface of endothelial cells, which is critical in angiogenesis (44), was found to be dependent upon the NH2-terminal heparan sulfate chains of perlecan (45). These patterns of perlecan expression clearly implicate a role in wound healing regulation and/or angiogenesis.
Cell and Growth Factor Binding
Perlecan is adhesive for fibroblasts and endothelial cells (28). Expression of perlecan in the mouse was coordinated with development of attachment competence by mouse embryos in vitro and in utero (7). Purified perlecan and laminin were found to promote attachment of immortalized rat chondrocytes in vitro (43). Perlecan is thought to modulate binding between the basement membrane structure and various cells, including smooth muscle cells and aortic endothelial cells, through a non-RGD cell binding region in one of the perlecan domains (possibly domain III in the mouse; amino acid sequence LPASFRGDKVTSY (SEQ ID NO: 1), as well as by GRGDSP (SEQ ID NO: 2), but not GRGESP (SEQ ID NO: 3)) and integrins β-I and α-5, β-3 (23). Alternatively, other investigators have found that the attachment of Rugli cells (a rat glioma derived cell line), mesenchymal, and epithelial derived tumor cell lines to mouse perlecan did not involve the protein core and was totally dependent on the presence of the heparan sulfate, although binding through the β-1 integrin of the cells was involved (4). Cell adhesion to perlecan was low compared to perlecan core alone (23). Human endothelial cell-derived perlecan was shown to bind endothelial cells in vitro with contributions from the heparan sulfate and from the protein core (46). The attachment of cells to the protein core of human perlecan further supports the implication of alternative cell binding pathways as the human homologue does not have the RGD sequence in domain III. Significant evidence for a role in cell binding further implicates perlecan in wound healing processes.
Proteoglycans such as perlecan, once thought to primarily serve as structural components of extracellular matrix, are now being focused on for their role in tissue and cell regulation, particularly angiogenesis and wound healing. Many growth factors, notably the fibroblast growth factor family (FGF) which now numbers 19 members, bind to heparin and heparan sulfate proteoglycans, and this binding has been shown to have a significant impact on the availability and activity of these growth factors. Importantly, perlecan has been shown to specifically bind to FGF-2 (also known as bFGF), which is critical in vascular development and wound healing (37). Perlecan was found to induce high affinity binding of FGF-2 both to cells deficient in heparan sulfate and to soluble FGF receptors. Further, in a rabbit ear model for in vivo angiogenesis, perlecan was a potent inducer of FGF-2-mediated neovascularization (2). It has been shown that FGF-2 binds to the heparan sulfate on domain I of perlecan (2) and that FGF-2 is released by biologically relevant enzymes such as plasmin, collagenase, and heparinases, which may have a role in the regulation of the growth factor activity (47). Mitogenic keratinocyte growth factor (FGF-7) was recently shown to bind to domains III and V of the perlecan protein core (31). Binding of FGF-2 to the heparan sulfate chain of perlecan is thought to involve three-way coordinated binding between FGF, heparan sulfate, and the FGF receptor, and to involve specific sites of sulfation (unpublished data). The FGF receptor is probably the FGF-binding protein that was recently reported to bind perlecan specifically in domain III (30).
Perlecan is also able to bind the growth factor granulocyte/macrophage-colony stimulating factor and present it to hematopoietic progenitor cells in a semi-solid colony assay (28). Bound growth factors can be released by enzymes, which are present during wound healing, such as the matrix metalloproteinases (47). Ligand binding itself can lead to internalization through a perlecan-mediated process, as has been shown for ligands such as lipoproteins (15). Binding of growth factors is clearly an important role of perlecan in wound healing and angiogenesis.
Breached basement membranes in vascular, corneal, and dermal tissues must respond quickly to the injury with repair. Such a rapid response suggests storage of needed biomolecules to effect repairs and remodeling. Cells involved in the wound healing and angiogenic process may have a ready and ample storage of FGF and other growth factors bound to perlecan in the basement membrane adjoining the wound. It has been demonstrated that enzymes which are turned on in the remodeling wound and angiogenesis, such as the matrix metalloproteinases (MMPs), stromelysin, rat collagenase, plasmin, urokinase, heparitinase I, and heparin, may modulate the bioavailability of the growth factors by degrading the protein core and removing the glycosaminoglycans (47). The MMPs are required in correct wound healing and in angiogenesis, and have been shown to bind to cells. The extracellular binding of MMPs could position the enzyme for directed proteolytic attack, for activation of other MMPs, and for regulation of other cell surface proteins. It has recently been demonstrated that heparin binds (5-10 nM) MMP-7, MMP-2, MMP-9, and MMP-13. This suggests that the MMPs may be positioned on the cell surface or retained in the ECM by perlecan heparan sulfate chains (48).
Growth and Wound Healing
The effectiveness of perlecan as an exogenously added promoter of growth and neovascularization was demonstrated with anti-sense perlecan knockouts in colon carcinoma cells (40). Growth of colon carcinoma cells was markedly attenuated upon obliteration of perlecan gene expression and these effects correlated with reduced responsiveness to, and affinity for, FGF-7. Exogenous perlecan effectively reconstituted the activity of FGF-7 in the perlecan-deficient cells. Moreover, soluble FGF-7 specifically bound immobilized perlecan in a heparan sulfate-independent manner. In both tumor xenografts induced by human colon carcinoma cells and tumor allografts induced by highly invasive mouse melanoma cells, perlecan suppression caused substantial inhibition of tumor growth and neovascularization. Thus, perlecan is a potent inducer of cell growth and angiogenesis in vivo and therapeutic interventions targeting this key modulator of tumor progression may improve wound healing.
In cells that were expressing antisense perlecan, responses to increasing concentrations of FGF-2 were dramatically reduced in comparison to wild-type or vector-transfected cells as measured by thymidine incorporation and rate of proliferation (3). Furthermore, receptor binding and affinity labeling of cells expressing antisense perlecan indicated that eliminating perlecan expression (by expressing antisense perlecan) results in reduced high-affinity FGF-2 binding. Both the binding and mitogenic response of cells expressing antisense perlecan to FGF-2 could be rescued by exogenous perlecan (3).
Poor wound healing in diabetics and in diabetes-related periodontitis may be related to perlecan expression. High levels of glucose can decrease perlecan expression in some cells, probably through transcriptional and post-transcriptional mechanism (27). Further, it has recently been demonstrated that FGF-2 bound to perlecan is protected from inactivation by non-enzymatic glycation, which occurs during the course of diabetes (35). The relationship between poor expression and protection from inactivation in diabetes has not been investigated.
Evidence indicates that in cases of poor surgical wound closure and healing by frank secondary intention, as results in many barrier membrane surgeries during oral and periodontal surgeries, for example, the presence of perlecan in the subepithelial matrix is associated with healing of the wound up to 14 days post surgery, at which time the healed wound can appear normal in terms of perlecan core protein. Subsequently, sulfation levels of the heparan sulfate chains increase over the ensuing year of healing (1). Another investigation, however, reported that perlecan is not detectable under epithelial cells migrating over connective tissue and only appears when the wound is covered and a new basement membrane is deposited (38). In wounds that are remodeling over a period of weeks, perlecan levels in the subepithelial connective tissues appear to remain relatively high (42). The functional significance of various perlecan expression patterns during healing is not yet known, but it is clear the biomolecule has an important and fundamental role in the healing process.
In summary, a complex interplay exists between 1) extracellular matrix proteins, including perlecan, 2) the cells considered key in establishing dermal integrity such as fibroblasts, endothelial cells, epithelial cells, and 3) growth factors such as the FGF s. It is clear, however, that perlecan has the potential to affect several aspects of wound healing and, to that end, is an excellent candidate for application or induction in various forms, with or without the addition of various growth factors to modulate wound healing.